Hacker News: docheinestages

New comment by docheinestages in "Claude Fable is relentlessly proactive"

docheinestages — Fri, 12 Jun 2026 11:37:46 +0000

I'm kind of on the fence about it and have a similar feeling. I don't mean to undermine the effort he has put in over all the years. That's definitely commendable. But I have strong suspicions that he's becoming an AI influencer, with his own AI focused newsletter, so chances are major AI companies are approaching him. And also to be honest, I see far too many posts making it to the front page. @dang I trust in the moderators keeping things neutral. Just in this thread alone there are a few comments that got heavily down voted for simply having a different opinion.

New comment by docheinestages in "A €0.01 bank transfer could compromise a banking AI agent"

docheinestages — Thu, 11 Jun 2026 01:34:56 +0000

I doubt there's a definitive and reliable fix, so long as AI agents are exposed to the "Lethal Trifecta" you wrote about. My guess is that it involves a series of fixes, similar to what the post describes, starting with low-hanging fruits like minimizing sensitive context and tool calls and breaking down agentic processes into a series of specialized agents with isolated capabilities and data. The long-term fixes in my opinon are remodelling this process based on zero trust principles and making it practically impossible for LLMs (or even a human who could take this role) to cause damage even if they wanted to.

New comment by docheinestages in "My automated doubt development process"

docheinestages — Sun, 07 Jun 2026 19:06:50 +0000

Then how come that process is not agentic and not well-described?

New comment by docheinestages in "My automated doubt development process"

docheinestages — Sun, 07 Jun 2026 19:03:56 +0000

Most writings about the spec-driven development I see start with a product requirements document that is assumed to be valid. But I doubt that's the case. If so, you would've written about it, and probably would've involved agents in the research that goes into it. My gut feeling tells me there's much more emphasis on implementing the feature than on questioning if it's relevant, feasible, and based on valid assumptions.

New comment by docheinestages in "LLMs are eroding my software engineering career and I don't know what to do"

docheinestages — Sun, 07 Jun 2026 18:44:22 +0000

In my experience, even the best frontier LLMs are very likely to make critical but subtle mistakes and false assumptions the more they're trying to one-shot the solution. One-shotting could be thought of as a broad term and varies depending on the use case. You have great results with LLMs because you did the job of finding the right documentation, and more importantly, those who wrote the documentation both had a deep understanding of the domain, and effectively compiled them into a coherent document. In other words, the more vetting, supervision, and research you do, the better the results. Of coruse, this doesn't mean doing the heavy lifting yourself. But the signal is key.

New comment by docheinestages in "LLMs are eroding my software engineering career and I don't know what to do"

docheinestages — Sun, 07 Jun 2026 16:00:48 +0000

LLMs can synthesize the domain knowledge so long as it's within their training data. At some point, blindly trusting the decisions they make becomes gambling.

New comment by docheinestages in "Sem: New primitive for code understanding – not LSPs, but entities on top of Git"

docheinestages — Sat, 06 Jun 2026 23:22:30 +0000

I doubt if this actually solves a real problem for humans or agents, especially in complex projects. It might help if the examples show scenarios where this tool and its commands could make a difference.

New comment by docheinestages in "Universal Memory Protocol – a shared format for agent memory"

docheinestages — Sat, 06 Jun 2026 23:04:12 +0000

Why should something like this make it to the front page?

New comment by docheinestages in "Programmers will document for Claude, but not for each other"

docheinestages — Fri, 05 Jun 2026 16:17:52 +0000

That could work, but there's still the chance that things could diverge if multiple people are working on a project and not everyone is as diligent as you. With LLM context windows continuously growing, agents should be able to scan the whole repository and even relevant repositories on the fly, provided they contain the truth and only what's necessary (i.e. minimal commentary).

New comment by docheinestages in "Stop Using Conventional Commits"

docheinestages — Fri, 05 Jun 2026 16:14:53 +0000

I think some structure in commit messages is helpful, but not to the point where it gets in the way of effectively reflecting what the commit contains, why it was done, and any comments for future reference, e.g. potential regressions.

New comment by docheinestages in "Programmers will document for Claude, but not for each other"

docheinestages — Fri, 05 Jun 2026 14:59:40 +0000

Generating tonnes of documentation is easy, but it can easily get outdated and so much that no one would read it. Ideally, code should be the single source of truth. Documentation should be generated dynamically and upon request to not go stale. The amount of detail and how far to dig in should be up to the end user.

New comment by docheinestages in "Mouseless – keyboard-driven control of macOS/Linux/Windows"

docheinestages — Fri, 05 Jun 2026 13:29:43 +0000

This is a helpful method for visually grounding LLMs to take actions on the screen such as clicking. For humans though, hell no.

New comment by docheinestages in "When AI Builds Itself: Our progress toward recursive self-improvement"

docheinestages — Thu, 04 Jun 2026 22:29:25 +0000

> We believe it would be good for the world to have the option to slow or temporarily pause frontier AI development to enable societal structures and alignment research to keep up with the advance of the technology.

Elon, is that you? [1]

[1] https://www.theguardian.com/technology/2023/mar/31/ai-resear...

New comment by docheinestages in "Gaussian Point Splatting"

docheinestages — Thu, 04 Jun 2026 12:14:09 +0000

Maybe use Tampermonkey?

New comment by docheinestages in "Odysseus – self-hosted AI workspace"

docheinestages — Sun, 31 May 2026 16:22:33 +0000

I really wonder if people who make such projects actually use them on a regular basis. I cannot imagine myself working with these UIs.

New comment by docheinestages in "The solution might be cancelling my AI subscription"

docheinestages — Sun, 31 May 2026 15:59:10 +0000

We're still in the phase where we're having our first reaction to the software development lifecycle with the help of AI. We're quickly starting to realize what AI is making cheap, and where the new bottlenecks are. How most people are currently using AI is rather naive and superficial. One-shotting only takes you so far.

New comment by docheinestages in "Is AI causing a repeat of frontend’s lost decade?"

docheinestages — Sat, 30 May 2026 13:31:08 +0000

The design specification can include everything you mentioned. The article argues that "JavaScript frameworks have deskilled frontend development in the last decade." That's what I can't agree with. We need the simplest way to achieve the desired outcome. Using higher-level means to do that doesn't mean deskilling.

New comment by docheinestages in "MCP is dead?"

docheinestages — Sat, 30 May 2026 11:22:33 +0000

Agreed. I think MCP should stay abstract in the sense tool-calling is. JSON-RPC could be one way to do it.

New comment by docheinestages in "CAPTCHAs can still detect AI agents"

docheinestages — Fri, 29 May 2026 17:16:57 +0000

I think it's just a game of cat and mouse. It might be easier to catch naive AI agents that are not fine-tuned for specific CAPTCHA tasks with human behavior, can't recognize new challenges, don't know when to stop and ask a human, and just want to brute force their way with limited or no specialized harness and tools available.

New comment by docheinestages in "We should be more tired than the model"

docheinestages — Fri, 29 May 2026 13:11:23 +0000

I think the best approach is active code review as the agent does small batches. Or letting it come up with a solution, testing if it passes or fails the desired outcome, then creating a separate fresh project and asking it to rewrite in small parts, and have it explain to you what and why it's doing to achieve each part.